SUMMARY We propose to screen a panel of human monoclonal antibodies (mAbs) to HIV-1 gp41 to identify the mAb that penetrates a model of the human blood brain barrier (BBB) most efficiently. We will test the use this mAb as a reagent for radioimmunotherapy (RIT) of the reservoir of HIV-1-infected cells in the central nervous system (CNS). During antiretroviral therapy (ART) which poorly cross BBB many HIV patients demonstrate HIV-associated neurocognitive disorders and neuroAIDS that continue to be a major public health problem. Migration of HIV-infected monocytes across the BBB causes neuroinflammation, which promotes dementia despite ART. Our in vitro and in vivo studies using human anti-gp41 mAb radiolabeled with the ?-particles- emitting radionuclide 213Bismuth revealed its ability to kill HIV-1-infected cells. We also showed the ability of this radiolabeled mAb to traverse a model of the human BBB in a manner that depends on its isoelectric point (pI) suggesting its therapeutic potential to treat HIV-related brain disorders. The in vitro BBB model is well established, tested and validated in a number of studies and several laboratories. This BBB model mimics the human BBB and uses astrocytes and brain microvascular endothelial cells which are cultured on both sides of a permeable membrane. We hypothesize that selection of new anti-gp41 mAbs with high pIs will increase the efficiency of radiolabeled mAb crossing BBB, and will kill HIV-1-infected cells on the CNS side of the BBB. Three specific aims are proposed: in Aim 1 we will screen a large panel of ~80 human anti-gp41 mAbs received from several sources including a panel of 16 mAbs produced in our laboratory. We will select mAbs with the highest isoelectric points that permit efficient passage across the BBB, high avidity binding to gp41 and broad binding to transfected cells expressing envelope proteins of viruses representing five different HIV-1 subtypes. In aim 2 we will concentrate on improving the binding capabilities of mAbs; the selected anti-gp41 mAbs in aim 1 will be used to produce bispecific antibodies with high pIs that target two different gp41 epitopes to increase their efficiency of targeting the broad spectrum of HIV-1-infected cells. We will also use molecular techniques to produce F(ab')2 fragments from selected anti-gp41 mAbs to investigate whether their smaller size increase their abilities to cross the model BBB. Aim 3 will test the efficiency of radiolabeled anti-gp41 mAbs to kill HIV-1-infected cells after penetration the BBB model. Given that transmission through the BBB model is time-dependent we will test mAbs labeled with three radionuclides with different half-life to select the most efficient construct. The radiolabeled anti-gp41 mAb can be used for future experiments in non-human primates infected with SHIV, and ultimately clinical trials.